Hello Everyone,   I am writing a SAS program where I get data from Compustat and CRSP in order to do an event study.  I merged my CRSP stock return data with my CRSP market index return data with no issues.  However, I am having issues merging my Compustat data with my CRSP data - for some reason, the match merge is not working, and it does not give me any error messages.  Can someone please help me?   Here is my SAS program - it uses remote connection to WRDS.  The issue is with the match merge at the end - in red font color.   /*Remote Sign-on to WRDS Server*/ %let wrds = wrds.wharton.upenn.edu 4016; options comamid=TCP remote=WRDS; signon username=_prompt_;   rsubmit; *Tell SAS where to look for data on WRDS; *Define directory for Compustat North America - Monthly Update;   libname compm '/wrds/comp/sasdata/nam';   data tem; set compm.fundq;  if indfmt="INDL";  if consol="C";  if popsrc="D";  if datafmt="STD";  * Get Compustat data for the Dow Jones Industrial Avg. Components for 2000 to 2015;  where 2000<=year(datadate)<=2015 and tic in ('AXP', 'AAPL', 'BA', 'CAT', 'CVX', 'CSCO',  'KO', 'DIS', 'DD', 'XOM', 'GE', 'GS', 'HD', 'IBM', 'INTC', 'JNJ', 'JPM', 'MCD', 'MMM',  'MRK', 'MSFT', 'NKE', 'PFE', 'PG', 'TRV', 'UNH', 'UTX', 'VZ', 'V', 'WMT');  Format cusip $8.;  Keep cusip tic conm datadate datafqtr rdq epsfxq; Run; * Download the dataset from WRDS to local library; Proc sort data=tem;  By cusip; Proc download data=tem out=cmpst_raw; Run; endrsubmit;   rsubmit; *Define directory for CRSP - Monthly Update; libname daily '/wrds/crsp/sasdata/a_stock'; Data tem2;  Set daily.dsf (keep = cusip date ret);  * Get CRSP Daily stock return data for the Dow Jones Industrial Avg. components from 2000 to 2015. CUSIP  is used since CRSP does not have ticker;   where 2000<=year(date)<=2015 and cusip in ('02581610','88579Y10','03783310','09702310','14912310',  '16676410','17275R10','89417000','38141G10','92826C83','59491810','93114210','91301710','91324P10',  '74271810','71708110','65410610','58933Y10','58013510','47816010','45920010','45814010','43707610');  format date yymmddn8.; Run; Proc sort data = tem2;  By date cusip;   * Download the dataset from WRDS to local library; Proc download data=tem2 out=crsp_stk; Run; endrsubmit;   rsubmit; libname ind '/wrds/crsp/sasdata/a_indexes'; * Get daily market index data for January 2000 to December 31, 2015; Data tem3;  set ind.dsic (keep=caldt vwretd ewretd);  Where '03Jan2000'd<=caldt<='31Dec2015'd;  Format date yymmddn8.;  Rename caldt=date; Run;   Proc sort data = tem3;  By date; Proc download data=tem3 out=crsp_indx; Run; endrsubmit;   Data crsp_stk_indx;  merge crsp_stk crsp_indx;  By date; Run; Data crsp_stk_indx;  set crsp_stk_indx;  where cusip ne " " and vwretd ne .; * Remove records that have no CUSIP or no Mkt return data; Run;   * Modify Compustat quarterly earnings file to calculate unexpected earnings based on previous earnings; Data cmpst_raw2;  Set cmpst_raw;  By cusip;  If epsfxq and rdq; * Keep records where earnings and report date for quarterly earnings are not missing;  ueps = epsfxq - lag(epsfxq); * UEPS is defined as the current earnings less prior qtr earnings;  If ueps < 0 then evntdum=1;  Else if ueps > 0 then evntdum=2;  Else evntdum = 0;  if first.cusip then ueps = 0;  If ueps; * Keep records where UEPS does not equal to zero;  keep cusip rdq epsfxq ueps evntdum; * Keep only CUSIP, Report date for quarterly earnings, earnings and UEPS; Run; * Create dates 90 days prior to event date (rdq) and 10 days after event date (rdq); data cmpst_int;  Format date yymmddn8.;  set cmpst_raw2;  do date = rdq - 90 to rdq + 10;     output;  end; keep cusip date rdq evntdum; run;   proc sort data=cmpst_int;  by cusip date; run;   Data crsp_stk_indx;  Set crsp_stk_indx (keep = cusip date ret vwretd evntdum); Proc sort data = crsp_stk_indx;  By cusip date; Run; Data crsp_stk_indx;  Retain date cusip ret vwretd;  set crsp_stk_indx; Run; Data crsp_cmpst;  Merge cmpst_int (in=a)  crsp_stk_indx (in=b);  By cusip date;  If a and b; Run;   Thanks!   ... View more

Hello,   I have an issue on how to figure out the SAS code for an event study that I am doing in SAS.  Specifically, I am running an event study on the 30 stocks of the Dow Jones Industrial Avg., from January 2000 to December 31, 2015.  The event study is related to the abnormal returns from unexpected earnings, or earnings surprises.  Therefore, I have done the following:   (1) I have created a data table with the daily stock and market index return data from CRSP for the 30 different stocks in the Dow Jones Industrial Avg. from January 3, 2000 to December 31, 2015.  The data table contains CUSIP, DATE, RET (Holding period return), and EWRETD (Equal Weighted Mkt Return incl. Dividend)   (2) I also created a table with the quarterly earnings data (eps diluted excl. extraordinary items) from Compustat for the 30 different stocks from January 2000 to December 2015.  The Compustat file has the following variables: CUSIP, EPSFXQ (eps for the quarter excl. extraordinary items), and RDQ (Report date for quarterly earnings - this will be my event date)   (3) To calculate unexpected earnings (UEPS), I am doing it in a very simple manner - specifically, I am taking the difference between the eps from quarters t and t-1. I know that this could pose issues if there is a strong seasonal variation in earnings, but for now, this is the easiest way.    (4) Therefore, I modified the Compustat data to include UEPS, and deleted all records where UEPS is equal to zero.  Therefore, I am only keeping the records where there is unexpected earnings (either greater than zero or less than zero).  This table effectively contains my Event Dates, which will be the RDQ for all records where UEPS is not equal to zero.   Now, comes the hard part, and this is where I need help with the SAS code.  For each CUSIP and RDQ (Event Date) from my modified Compustat data table, I want to create dates that are 90 days before the Event Date, and 10 days after the Event Date -- therefore, effectively, I want to have 101 days for each Event Date and CUSIP combination.  Then, I want to merge this data with my CRSP stock and market index return data.  Specifically, here are some sample data, and what I want to see as the final outcome:   (A) CRSP Stock and Mkt Return Daily Data   DATE                CUSIP                  RET                 EWRETD 20000103         19121610             -0.032189         0.002878 20000103         25468710              0.021368         0.002878 20000103         26353410             -0.013283         0.002878 20000104         19121610             -0.008869        -0.017466 20000104         25468710              0.062762        -0.017466 20000104         26353410              0.000000        -0.017466   (B) Compustat Modified Data   CUSIP                RDQ                     EPSFXQ             UEPS 19121610           20000719             0.37                     0.39 19121610           20001020             0.43                     0.06 19121610           20010131             0.10                     -0.33 25468710           20000803             0.21                     0.13 25468710           20001109             0.11                    -0.10   (C) I want to create a file that has the following resulting data&colon;   CUSIP                RDQ                   DATE                 RET                EWRETD 19121610           20000719          20000419          0.0310              0.00145 19121610           20000719          20000420          0.0025             -0.02432 19121610           20000719          20000421         -0.0119              0.02501 19121610           20000719          20000422          0.0222              0.00541   Therefore, for each CUSIP and RDQ combination, I want to have daily stock return and market index return for 90 days before the RDQ (Event Date) and 10 days after the RDQ (Event Date).   Can someone please help me with the SAS code to do this?  Since I am a beginner in SAS, it would be great if someone could show me the code without using PROC SQL.   Thanks everyone!                     ... View more